The aim was to test Todd-Hewitt broths (TH-broths) for the ability to propagate pneumococci and thereafter to evaluate the serotyping result obtained by the Pneumotest-Latex kit (SSI). TH-broths from four different producers (Oxoid, Sigma, Difco, and SSI) were tested and compared separately and with Serum broth (SSI). Twenty-three pneumococcal strains (different serotypes) were inoculated into the broths with start inoculums of 101, 10 3, and 106 CFU/ml. After incubation, overnight viable bacterial counts and visible growth were recorded. All pneumococci were serotyped with the Pneumotest-Latex kit. After incubation, the bacterial counts in all TH-broths were within the range of log 4.65-log 7.76 CFU/ml, while Serum broth showed an average growth ranging from log 8.05-log 8.90 CFU/ml. Comparing the growth of the four TH-broths showed no significant differences. In general, Serum broth had a more pronounced visual growth than each of the four TH-broths. Serotyping with Serum broth showed in general positive and correct latex typing results for all serotypes and initial inoculum, while the outcome of the TH-broths showed some false negative results depending on inoculum and serotype. Overall the Serum broth was found to be superior to the four TH-broths tested both with regard to CFU/ml and when used with the Pneumotest-Latex kit. However, if the Pneumotest-latex kit is only used on broths with visible growth as stated in the instruction manual, then the differences between the performances of the broths from the different producers was not significant and all broths could be used for Pneumotest-Latex typing.